Finally, some recognition that transparency is key to trust in AI - we can't just take "it works" for an answer. Explainability should be a fundamental aspect of model design, not an afterthought. https://news.mit.edu/2026/improving-ai-models-ability-explain-predictions-0309